PROJECT SUMMARY While late onset sporadic Alzheimer's disease (AD) is usually announced with amnesia, the aggregated proteins ?-amyloid (A?) and tau probably deposit in the brain for many years before symptom onset. This process occurs in the brain's episodic memory system, on a background of normal aging. Increasing evidence points to the spread of the tau protein out of the medial temporal lobe and into neocortical brain regions as crucial in the transition from normal aging to AD, possibly driven by A? and patterns of neural activity and connectivity. In this project we will specifically examine two subsystems of the episodic memory system, an anterior temporal (AT) system originating in lateral entorhinal cortex (LEC) specialized for object memory, and a posteromedial system (PM) system originating in medial entorhinal cortex (MEC) specialized for spatial memory. This is important for differentiating aging and AD because tau deposition begins in the LEC in older brains, while A? deposits in the PM system. The application builds upon on a longitudinal cohort consisting of almost 200 cognitively normal older people who have previously had baseline amyloid PET scanning with [11C]PIB, longitudinal structural MRI exams, and some of whom have had tau-PET imaging with [18F]flortaucipir. For this project, 120 participants will undergo a baseline examination of PIB-PET, flortaucipir- PET, structural MRI, and neuropsychological testing of memory and other cognitive abilities; these procedures will be repeated 1.5 and 3 years later. The baseline examination will also include a functional MRI experiment in which participants encode novel objects and scenes to define the AT and PM episodic memory systems. Neural activity will be examined using directed functional connectivity (directed-FC), an analytic approach employing Granger causality with assessment of neural activity directed from one region to another. This directed-FC will model the spread of tau through memory systems over the ensuing 3 years. Aim 1 will examine the overall pattern of tau spread in relation to the presence of A?, longitudinal cortical atrophy, and cognitive change. We hypothesize that A? will speed tau spread, which in turn will be associated with atrophy and memory decline. Aim 2 will examine tau spread through the AT and PM systems. We hypothesize that tau spreads predominantly in the AT system and, as such, reflects an enhancement of the processes that begin in the aged brain and not an anatomical or functional new condition. Aim 3 will use directed-FC to predict the spread of tau over time, with the hypothesis that brain regions most strongly connected to the entorhinal cortex will show the most rapid spread of tau pathology and this will be strongest in the AT system. How tau spreads through these systems and how A? and neural connectivity may drive this spread could help to differentiate the earliest stages of AD from normal aging, identify normal individuals at highest risk of progression, and provide new approaches to the selection of individuals for clinical trials.